Redis

• an in-memory key-value store that can be used as a database, cache or message broker.
• Redis began as a caching database, but it has since evolved into a primary database. Many applications built today use Redis as a primary database. However, most Redis service providers support Redis as a cache but not as a primary database.
• you can use Redis as both an in-memory cache and a primary database in a single system, thus eliminating the complexity and latency of two separate systems

Redis as a Cache vs Redis as a Primary Database

• scalability (Scaling)

• fast performance » real-time analytics, chat/messaging, media streaming, and pub/sub apps.

• https://redis.com/blog/redis-cache-vs-redis-primary-database-in-90-seconds/#:~:text=Redis%20began%20as%20a%20caching,not%20as%20a%20primary%20database.

Redis VS Memcached

• Memcached에 비해 다양한 기능을 제공하며 싱글 스레드 방식으로 동작하여 가볍기 때문입니다.

Spring Data Redis (What I used)

https://docs.spring.io/spring-data/data-redis/docs/current/reference/html/

application.yml

1
2
3
4
5
6
7
8
9
10
11
12
13
14

spring:
  cloud:
    cache:
      type: redis
      cache-names:
        - ***
        - ***
        - ***
    redis:
      host: ***
      lettuce:
        pool:
          max-active: **
          min-idle: **

lettuce

• a Redis client of Java that is fully non-blocking.
• It supports both synchronous and asynchronous communication.
• scalable
• thread-safe
• https://www.baeldung.com/java-redis-lettuce

Jedis VS Lettuce

• commonality : Redis에 접근하기 위해서 사용하는 Spring Data Redis Library (성능과 사용성을 위해 만들어진 클라이언트)
• Jedis : deprecated
  • 동기 방식으로 작동하여 Blocking 이슈가 발생 가능하다는 단점이 있습니다.
• Lettuce
  • Lettuce 역시 클라이언트 라이브러리이며 동기, 비동기 방식을 둘 다 지원하여 non-blocking하게 요청을 처리할 수 있고 확장성이 뛰어나다는 장점이 있습니다. 하지만 사용성이 Jedis에 비해 어렵다는 단점이 있습니다.

** 빠르게 쉽게 개발할 필요성보다 추후 변경 사항 및 튜닝 등을 대비하여 확장성이 높으며 비동기 처리를 통해 안정적으로 요청을 처리할 수 있는 Lettuce가 더 적합

• https://abbo.tistory.com/107
• https://velog.io/@gale4739/Spring-Boot-Redis-%EC%A0%81%EC%9A%A9%EA%B8%B0-With-lettuce
• https://redis.com/blog/jedis-vs-lettuce-an-exploration/#:~:text=What%20is%20Lettuce%3F,Sentinel%2C%20Pipelining%2C%20and%20codecs.

• non-blocking VS blocking
  • https://1-7171771.tistory.com/131
• pooling
  • max-acitve : Maximum active connections to Redis instance
  • maxidle : the max number of connections that can be idle in the pool without being quickly closed (the default value is 8)
  • minIdle : This is the number of connections that are ready for immediate use.
  • https://medium.com/geekculture/the-pooling-of-connections-in-redis-e8188335bf64#:~:text=maxIdle%3A%20This%20is%20the%20max,a%20short%20period%20of%20time.

max-active VS max-idle

• max-active : the maximum number of active connections that can be allocated from this pool at the same time.
• max-idle : The maximum number of connections that should be kept in the pool at all times.
• https://stackoverflow.com/questions/9451818/what-is-the-difference-between-maxactive-vs-maxidle-for-tomcat-connection-pools

how to use Redis-cli

1
2
3
4
5

brew install redis ( default port : 6379)
redis-cli --raw -h ... -p ... (--raw : 한글 안깨져서 나옴)
>> ping (접속 테스트)
>> keys * 
>> get (key name)

Caching

• 한번 처리한 데이터를 임시로 저장소에 저장하는 것으로, 이 임시 데이터를 동일하거나 유사 요청이 왔을 경우 저장소에서 바로 읽어와서 응답을 하여 성능 및 응답속도 향상을 위한 기술

@Cachable vs @CacheEvict

• @Cacheable
  • 캐시가 있으면 캐시의 정보를 가져오고, 없으면 등록한다.
• @CacheEvict
  • 캐시삭제
  • 먼저 Evict 에 대해서는 Cache 자체를 지운다는 의미보다는, 메모리가 부족해서 더 이상 캐시할 수 없을 때, 메모리를 확보하기 위해서, 기존 캐시된 데이터를 지우는 것을 의미합니다.
  • https://charsyam.wordpress.com/2022/04/18/%EC%9E%85-%EA%B0%9C%EB%B0%9C-spring-%EC%9D%98-cacheevict-%EC%97%90%EC%84%9C-allentriestrue-%EB%8A%94-redis%EC%97%90%EC%84%9C-%EC%96%B4%EB%96%BB%EA%B2%8C-%EB%8F%99%EC%9E%91%ED%95%98%EA%B2%8C/

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

@Cacheable(cacheNames = CacheNames.DISCLOSURE)

@Cacheable(cacheNames = CacheNames.ADD_ATTR, key = “#root.methodName”, condition = “#result != null”)

@Cacheable(cacheNames = CacheNames.SUPPLIER, key = “#supplierCode”)

@Cacheable(value = “ownerCode”, key = “#result  == null”)
public String getOwnerCode() {...}

@Cacheable(cacheNames = CacheNames.PROMO, key = “#root.methodName”, unless =  “#result.isEmpty()”)

@CacheEvict(cacheNames = CacheNames.DISCLOSURE)

@CacheEvict(cacheNames = ”supplier”, key = “#request.supplierCode”)
public void changeSupplier(@NotNull SupplierModifyRequest request) {...}

• https://yonguri.tistory.com/82

Redis Sharding

• Sharding, also known as partitioning, is splitting the data up by key; While replication, also known as mirroring, is to copy all data. Sharding is useful to increase performance, reducing the hit and memory load on any one resource. Replication is useful for getting a high availability of reads.
• Sharding is used to scale Writes while Replication is used to scale Reads.
  • https://stackoverflow.com/questions/2139443/redis-replication-and-redis-sharding-cluster-difference#:~:text=Sharding%2C%20also%20known%20as%20partitioning,a%20high%20availability%20of%20reads.

Redis Redis Çluster

• Redis scales horizontally with a deployment topology called Redis Cluster.
  • https://redis.io/docs/management/scaling/
• Advantages
  • High Performance
  • High Availability
  • Horizontal & Vertical Scalability
  • Native Solution
• https://scalegrid.io/blog/intro-to-redis-cluster-sharding-advantages-limitations-deploying-and-client-connections/